One of the workshops I attended at the Web 2.0 Conference last week was about Mash-ups and asked the question: Where's the Business Model? It was moderated by Dave McClure from SimplyHired.com and featured Jeffrey McManus from Yahoo! (no relation), Paul Rademacher from Google (recently hired due to his outstanding work on the quintessential mash-up HousingMaps.com) and Adam Trachtenberg from eBay.
Paul Rademacher talked about how Housing Maps, which is a mash-up of Google Maps and craigslist, came out of a real world need for him and grew from that. The lesson here is that usefulness is a key driver for success in the Web mash-ups world. He also mentioned that Housing Maps isn't necessarily a pure API mash-up, because the craigslist data is taken from their RSS feeds and then he stores the content on his own server. So mash-ups are still painful to implement, despite all the hype around using APIs to deliver them. Paul also said he that reached 900,000 visitors to Housing Maps without any marketing effort, which is an indication of how mash-ups can be low-cost to both produce and market.
Because mash-up services usually don't own or control the data, the way most mash-ups compete is on having a better or more flexible User Interface, or adding extra data. This is a crucial point about the business models for mash-ups - because the data is open, what matters is what you do with the data and how you present or deliver it. In a way this represents what the ideal open Web 2.0 model is all about - free the data and innovate on top of that.
Further, the more data sets you try to mash up - the more "periphery opportunities" (as Paul called them) there will be. Dave McClure talked about his site, SimplyHired.com, integrates Google Maps, DMB for company information, LinkedIn for career and social networking data, and payscale.com. Dave called this a "lightweight integration" and said there was still "a lot of scraping going on" rather than using formal APIs.
Adam talked about a mash-up between eBay Motors and Google Maps, which uses the APIs of those two companies. It enables users to find vehicles for sale in a location near them. The data in this mash-up is not stored on the host's server (unlike Housing Maps) but is served up in real time. Adam talked about how this mash-up gave the user experiences they wouldn't normally get on eBay - primarily the visual mapping experience. But the app still does a lot of things users can do on the eBay website, for example adding a watchlist onto their eBay profile. There is also an eBay affiliate program, which is a possible source of revenue.
Jeffrey McManus discussed how Yahoo! Maps was used to track Hurricane Rita recently, illustrated in this screenshot. It transformed weather service data into a visual and easy way to see the path of Rita. This was done by caching the data onto their server and then presenting it on a Yahoo! Maps UI.
Those are just some of the examples of mash-ups discussed in the workshop. The necessary components of a mash-up according to the panelists are AJAX or a similar client-side technology, APIs in the backend (or RSS and/or scraping if need be - although some people argue scraping isn't truly Web 2.0), and a lot of work on the UI and data sets by the developer(s). Paul mentioned there is still a lot of work to be done with the technologies - e.g. authentication. So it's early days yet for mash-ups on the Web.
As for mash-up business models, some of the suggestions were: advertising, lead generation and/or affiliates, transactional, subscription. I'm sure we'll see a lot more business innovation, as well as technological innovation, with mash-ups over the next year or so.